PCA-改进RPROP方法的BP算法在音乐信号分类中的应用
|
摘要
针对标准的BP神经网络仅从预测误差负梯度方向修正权值和阈值,学习过程收敛缓慢,并且容易陷入局部最小值,导致泛化能力不足的问题,提出了一种基于学习经验变学习速率改进的RPROP方法作为BP神经网络权值和阈值更新方法,并与主成分分析法(Principal Component Analysis,PCA)相结合,形成了PCA-改进神经网络算法。同时,采用Matlab软件对四类音乐信号进行分类实验。实验结果表明,改进算法比标准算法的稳定识别率提高2.6%,当稳定识别率达到90%时,用时节省75%,表明该算法可以加快网络的收敛过程,提高泛化能力。
The standard BP neural network corrects the weights and thresholds from the negative direction of the prediction error,which causes the learning process to converge slowly and easily fall into the local minimum,resulting in insufficient generalization ability.In order to solve this problem,an improved RPROP method based on learning experience variable learning rate is proposed as BP neural network weight and threshold updating method,and combined with principal component analysis(PCA)to form PC A-modified neural network algorithm.By classifying the four types of music signals in Matlab software,the final classification results show that the improved algorithm is 2.6% higher than the standard algorithm,and it can save 75% of time when the stable recognition rate reaches 90%.The algorithm can speed up the network convergence process,and improve the generalization ability.
The standard BP neural network corrects the weights and thresholds from the negative direction of the prediction error,which causes the learning process to converge slowly and easily fall into the local minimum,resulting in insufficient generalization ability.In order to solve this problem,an improved RPROP method based on learning experience variable learning rate is proposed as BP neural network weight and threshold updating method,and combined with principal component analysis(PCA)to form PC A-modified neural network algorithm.By classifying the four types of music signals in Matlab software,the final classification results show that the improved algorithm is 2.6% higher than the standard algorithm,and it can save 75% of time when the stable recognition rate reaches 90%.The algorithm can speed up the network convergence process,and improve the generalization ability.
引文
[1]高广银,丁勇,姜枫,等.基于BP神经网络的停车诱导泊位预测[J].计算机系统应用,2017(1):236-239.
[2]Yu F,Xu X Z.A short-term load forecasting model of natural gas based on optimized genetic algorithm and improved BPneural network[J].Applied Energy,2014,134:102-113.
[3]翟俊海,苗青,李塔,等.概率神经网络样例选择算法[J].小型微型计算机系统,2015,36(4):787-791.
[4]刘加存,苑增福,梅其祥.基于分数阶滤波器的高泛化性神经网络建模[J].测控技术,2017(12):57-62.
[5]朱江淼,宋文峰,高源,等.基于改进型BP神经网络的氢原子钟钟差预测[J].仪器仪表学报,2016,37(2):454-460.
[6]Zweiri Y H,Whidborne J F,Seneviratne L D.A three-term backpropagation algorithm[J].Neurocomputing,2003,50(1):305-318.
[7]赵燕伟,任设东,陈尉刚,等.基于改进BP神经网络的可拓分类器构建[J].计算机集成制造系统,2015,21(10):2807-2815.
[8]魏微,高谦.改进的BP神经网络模型预测充填体强度[J].哈尔滨工业大学学报,2013,45(6):90-95.
[9]张凌波,宰娜,顾幸生.基于改进教学算法优化BP神经网络的催化剂碳含量预测模型[J].控制与决策,2016,31(9):1723-1728.
[10]Riedmiller M,Braun H.A direct adaptive method for faster backpropagation learning:the RPROP algorithm[C]//IEEE International Conference on Neural Networks.1993:586-591.
[11]杨存祥,朱琛,解豪杰.基于RPROP神经网络算法的异步电动机故障诊断[J].电力自动化设备,2012,32(1):80-83.
[12]李康顺,李凯,张文生.一种基于改进BP神经网络的PCA人脸识别算法[J].计算机应用与软件,2014(1):158-161.
[13]章剑光,周浩,盛晔.基于RPROP神经网络算法的主变DGA故障诊断模型[J].电力系统自动化,2004,28(14):63-66.
[14]Hecht-Nielsen R.Theory of the backpropagation neural network[C]//International 1989 Joint Conference on Neural Networks.1989.
[2]Yu F,Xu X Z.A short-term load forecasting model of natural gas based on optimized genetic algorithm and improved BPneural network[J].Applied Energy,2014,134:102-113.
[3]翟俊海,苗青,李塔,等.概率神经网络样例选择算法[J].小型微型计算机系统,2015,36(4):787-791.
[4]刘加存,苑增福,梅其祥.基于分数阶滤波器的高泛化性神经网络建模[J].测控技术,2017(12):57-62.
[5]朱江淼,宋文峰,高源,等.基于改进型BP神经网络的氢原子钟钟差预测[J].仪器仪表学报,2016,37(2):454-460.
[6]Zweiri Y H,Whidborne J F,Seneviratne L D.A three-term backpropagation algorithm[J].Neurocomputing,2003,50(1):305-318.
[7]赵燕伟,任设东,陈尉刚,等.基于改进BP神经网络的可拓分类器构建[J].计算机集成制造系统,2015,21(10):2807-2815.
[8]魏微,高谦.改进的BP神经网络模型预测充填体强度[J].哈尔滨工业大学学报,2013,45(6):90-95.
[9]张凌波,宰娜,顾幸生.基于改进教学算法优化BP神经网络的催化剂碳含量预测模型[J].控制与决策,2016,31(9):1723-1728.
[10]Riedmiller M,Braun H.A direct adaptive method for faster backpropagation learning:the RPROP algorithm[C]//IEEE International Conference on Neural Networks.1993:586-591.
[11]杨存祥,朱琛,解豪杰.基于RPROP神经网络算法的异步电动机故障诊断[J].电力自动化设备,2012,32(1):80-83.
[12]李康顺,李凯,张文生.一种基于改进BP神经网络的PCA人脸识别算法[J].计算机应用与软件,2014(1):158-161.
[13]章剑光,周浩,盛晔.基于RPROP神经网络算法的主变DGA故障诊断模型[J].电力系统自动化,2004,28(14):63-66.
[14]Hecht-Nielsen R.Theory of the backpropagation neural network[C]//International 1989 Joint Conference on Neural Networks.1989.